Once upon a time, I was an amateur bard performing songs and stories. But now, asthma and sinusitis have given me a nasal and breathless delivery. I never could manage reading dialog and action very well; I was better at a folktale delivery where one didn't have to indicate change of speaker.
Option 2 is to use NaturalReader software to create Mp3s artificially. The pronunciation editor is flaky-- rather than let me type phonics, I have to try various misspellings and hope it gets one of them right-- and sometimes it gets the emphasis in a word or sentence incorrect. But on the whole, it does a remarkably good job.
Either way, I have to do a certain amount of editing and processing to get a passable recording, snipping out my stammers and pops or beating on Heather's wonky pronunciations and pacing to the extent that I can.
So here's recordings for Chapter 44 of Love Her and Despair, broken into two chunks since it's a long chapter:
Live recording: Part One | Part Two
Text-to-Speech: Part One | Part Two
Which is the easiest to listen to?
Auronlu's: asthma, sinusitis, and Shatneresque voice acting notwithstanding, I prefer a human
Computer: Heather's got a yummy voice, even with the occasional timing and pronunciation glitches
Advent Children's godawful English dub